Symmetry-Breaking for Formation of Rectangular CdSe Two-Dimensional Nanocrystals in Zinc-Blende Structure

Article abstract of DOI:10.1021/jacs.7b04855

Formation of CdSe nanocrystals with two-dimensional quantum confinement (CdSe 2D nanocrystals) was studied with preformed CdSe nanocrystals in the size range between 1.7 and 2.2 nm as seeds. Specifically, the 2D CdSe nanocrystals were encased with six {100} facets of the zinc-blende (face-center-cubic) structure, that is, 1.5 nm in thickness with quite large atomically flat {100} basal planes (～8 nm width and X ≈ 45 nm length). Symmetry breaking between the thickness and lateral directions occurred in the early stage by rapid formation of single-dot intermediates with flat yet polar {100} basal planes and the desired thickness from the seeds through intraparticle ripening. Two single-dot intermediates fused together through their reactive side facets - mostly the nonpolar {110} ones - to form 2D embryos with the same thickness. Such oriented attachment continued selectively onto the reactive side facets of the 2D embryos. Simultaneously, intraparticle ripening occurred slowly on the side facets of the 2D nanocrystals, which converted unstable side facets gradually to four stable {100} ones. When ～3 stable {100} side facets were developed, oriented attachment would continue on the remaining active one, which would result in the second symmetry breaking between two lateral directions. Cadmium acetate assisted both formation of single-dot intermediates and oriented attachment. Cadmium alkanoates with a long hydrocarbon chain selectively stabilized polar {100} facets on the nanocrystals including single-dot intermediates and shuttled insoluble acetate to the reactive surface of the nanocrystals.

Full text of DOI:10.1021/jacs.7b04855

Article
Symmetry-Breaking for Formation of Rectangular CdSe
Two-Dimensional Nanocrystals in Zinc-Blende Structure
Yiya Chen, Dongdong Chen, Zheng Li, and Xiaogang Peng
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.7b04855 • Publication Date (Web): 30 Jun 2017
Downloaded from http://pubs.acs.org on July 1, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of the American Chemical Society is published by the American Chemical
Society. 1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the
course of their duties.

Page 1 of 32
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Symmetry-Breaking for Formation of Rectangular CdSe Two-Dimensional Nanocrystals in
Zinc-Blende Structure
Yiya Chen1, Dongdong Chen1, Zheng Li2*, Xiaogang Peng1*
1Center for Chemistry of Novel & High-Performance Materials, Department of Chemistry, Zhejiang University,
Hangzhou 310027, China
2State Key Laboratory of Chemo/Bio-Sensing and Chemometrics, College of Chemistry and Chemical
Engineering, Hunan University, Changsha 410082, China
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
-------------------------------------------------------------------------------------------------------------------------
Abstract: Formation of CdSe nanocrystals with two-dimensional quantum confinement (CdSe 2D
nanocrystals) was studied with pre-formed CdSe nanocrystals in the size range between 1.7-2.2 nm as
seeds. Specifically, the 2D CdSe nanocrystals were encased with six {100} facets of the zinc-blende
(face-center-cubic) structure, i.e. 1.5 nm in thickness with quite large atomically-flat {100} basal
planes (~8 nm width X ~45 nm length). Symmetry breaking between the thickness and lateral directions
occurred in the early stage by rapid formation of single-dot intermediates with flat yet polar {100}
basal planes and the desired thickness from the seeds through intra-particle ripening. Two single-dot
intermediates fused together through their reactive side facets—mostly the nonpolar {110} ones—to
form 2D embryos with the same thickness. Such oriented attachment continued selectively onto the
reactive side facets of the 2D embryos. Simultaneously, intra-particle ripening occurred slowly on the
side facets of the 2D nanocrystals, which converted unstable side facets gradually to four stable {100}
ones. When ~3 stable {100} side facets were developed, oriented attachment would continue on the
remaining active one, resulting in the second symmetry breaking between two lateral directions.
Cadmium acetate assisted both formation of single-dot intermediates and oriented attachment.
Cadmium alkanoates with a long hydrocarbon chain selectively stabilized polar {100} facets on the
nanocrystals including single-dot intermediates and shuttled insoluble acetate to the reactive surface
of the nanocrystals.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 32
1
2
3
4
5
Introduction
6
7
8
9
Colloidal semiconductor nanocrystals are attracting tremendous attention because of their dimension-
dependent optical properties. 1 With their three-dimensions in the quantum-confinement size regime,
dot-shaped semiconductor nanocrystals are known as quantum dots or zero-dimensional (0D)
nanocrystals.1 Rod-shaped nanocrystals with their long axis as the only dimension beyond the exciton
size of the corresponding bulk materials are considered as quantum rods, i.e., one-dimensional (1D)
nanocrystals. 2 Colloidal nanocrystals with only one dimension smaller than the quantum-confinement
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3
threshold of the bulk exciton are established as two-dimensional (2D) nanocrystals. Synthesis of
4,5
high-quality colloidal 0D nanocrystals have been advanced rapidly since the early 1990s,
and
monodisperse quantum rods became available in 2000. 2 Rapid progress on 0D and 1D nanocrystals is
greatly benefited by active studies of their corresponding growth mechanisms, which are reasonably
unified at least for the most studied systems, i.e., CdSe and other II-VI nanocrystals. 6-18 In comparison,
2D nanocrystals are under-developed. For the mostly developed II-VI 2D nanocrystals, the first report
3
on CdSe 2D nanocrystals in wurtzite structure appeared in 2006 and those in zinc-blende structure
were reported in 2008. 19 The predicament on colloidal synthesis of 2D nanocrystals is related to poor
20,21
understanding of their formation mechanism, especially the necessary symmetry breaking.
Symmetry breaking here refers to the preferred growth along one of symmetry-equivalent and isotropic
directions. For instance, II-VI semiconductors are typically in either wurtzite (hexagonal) or zinc-
blende (face-center-cubic) phases, whose symmetry does not support formation of 2D nanocrystals
with atomically flat basal planes with significantly different dimensions on thickness, width, and length.
Though it is challenging to synthesize II-VI 2D nanocrystals, their optical properties are qualitatively
ACS Paragon Plus Environment

Page 3 of 32
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
different from those of the corresponding 0D and 1D nanocrystals. For instance, the homogeneous
peak width of photoluminescence (PL) of 0D and 1D nanocrystals is quite similar, but that of the
corresponding 2D nanocrystals with similar PL peak position can be reduced by one-half. 22,23 The PL
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
decay dynamics of the 2D nanocrystals is also a magnitude faster than that of the corresponding 0D
24,25
and 1D nanocrystals.
The Stoke’s shift of all 2D cadmium chalcogenide nanocrystals is nearly
zero 26-28 while that of the corresponding 0D and 1D ones is quite large. 2,5 All these intriguing optical
features are believed to be associated with their atomically-flat basal planes, which are perpendicular
to the quantum-confinement direction of the 2D nanocrystals. For II-VI nanocrystals, only those tiny
“magic-size clusters” are confirmed to possess atomically-flat facets. 29-31
As the most developed system, formation mechanisms of CdSe 2D nanocrystals in wurtzite and zinc-
blende structures are suggested to be completely different from each other. Wurtzite CdSe 2D
nanocrystals are proposed to be formed by assembly/attachment of pre-formed magic-size clusters
within the soft templates of alkylamine-alkanoates layer structures. 22,32 Conversely, formation of zinc-
blende CdSe 2D nanocrystals has been suggested by lateral extension of properly pre-formed seeds—
23,27,28,33
selective addition of the monomers onto the sides of seeds.
While symmetry-breaking is
suggested through the layer-structures of the templates for the wurtzite case, it is largely an un-
answered question for the zinc-blende case. It is interesting to notice that both basal planes and lateral
side facets of zinc-blende CdSe 2D nanocrystals can be the same, i.e., {100} facets of the face-center-
cubic (fcc) structure, 23,33 which makes the symmetry-breaking more bewildering.
The Weller’s group has studied formation of colloidal PbS 2D nanocrystals (nanosheets), which is also
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 32
1
2
3
4
5
6
7
8
9
34
in a face-center-cubic structure (rock salt). They have suggested oriented-attachment —the crystal
growth through fusion of primary nanocrystals with their lattices aligned to each other—as the
formation mechanism, and the symmetry-breaking is achieved by direction of attachment, namely by
fusing their reactive {110} facets of adjacent nanocrystals. 35
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Though distinguishable mechanisms are proposed for zinc-blende CdSe and rock salt PbS 2D
nanocrystals, there are quite some similarities between these two systems. Firstly, both zinc-blende
CdSe and rock salt PbS are face-center-cubic crystal structure. Secondly, synthesis of relatively thick
CdSe 2D nanocrystals—only available for zinc-blend ones yet—and PbS 2D nanosheets both require
23,35
pre-formation of nanocrystal seeds,
which are significantly larger than the thickness of the final
2D nanocrystals (see below). Thirdly, in both cases, some adventitious reagents are required to initiate
formation of the 2D nanocrystals from the seeds, namely chlorine-containing organics for the PbS
system and metal acetates for the CdSe system. Noticeably, acetates were found to be the key reagents
to initiate oriented attachment of copious types of inorganic nanocrystals by reducing their surface
passivation. 36-40
The analysis above implies that oriented-attachment might be a plausible mechanism for formation of
zinc-blende CdSe 2D nanocrystals, instead of simple lateral extension by selective monomer addition.
This work intends to systematically investigate this alternative mechanism. Different from previous
studies, the pre-formed CdSe nanocrystal seeds with a range of sizes are purified from the residual
41
precursors and side products using an established procedure. Emphasis are on identification of
symmetry breaking between three directions, given the dimensions of the thickness, width, and length
ACS Paragon Plus Environment

Page 5 of 32
Journal of the American Chemical Society
1
2
3
4
being substantially different from each other for the CdSe 2D nanocrystals.
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Results and discussion
Design of experimental system. Figure 1a illustrates the general scheme of the experiments. CdSe
quantum dots with their diameter varied between 1.7 and 2.2 nm—with 1.8 nm as the default size (see
Figure S1, Supporting Information)—were applied as the seeds. These nanocrystal seeds were
synthesized with cadmium n-alkanoates and elemental Se as the precursors in octadacene (ODE). The
nanocrystal seeds were purified using a recently developed procedure that quantitatively removed the
41
unreacted precursors and free fatty acids. The purified seeds were injected into a growth solution
with
designated
and
compositions
temperature. During
the growth, aliquots
with a known volume
were taken at different
time intervals from the
reaction solution and
used for monitoring the
reaction. To ensure
dissolution of the
reaction mixture for
Figure 1. (a) Schematic illustration of the synthetic protocol for the CdSe 2D
nanocrystals. (b) UV-Vis absorption spectra of purified CdSe seeds and the as-
synthesized 2D nanocrystals. (c) Typical TEM image of the CdSe 2D
nanocrystals. (d) Side view of the atomic model for the CdSe 2D nanocrystals.
(e) HRTEM image of a CdSe 2D nanocrystal with the corresponding FFT
pattern (inset).
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 32
1
2
3
4
5
6
7
8
9
quantitative measurements, the aliquots were diluted with a known amount of toluene with ~0.2% oleic
acid (or octanoic acid), which should convert the surface ligands into entropic type to promote
solubility. 42
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The scheme in Figure 1a was modified from a synthetic procedure originally developed by the
Dubertret group. 19 Without purification of the seeds, the original procedure introduced metal acetates
into the solution of the pre-formed seeds, which was confirmed to yield zinc-blende CdSe 2D
nanocrystals with alternating stacking of 6 monolayers of Cd cations and 5 monolayers of Se anions,
namely 5.5 monolayers of CdSe unites along the thickness direction and 1.5 nm in thickness from the
top to bottom basal planes (Figure S2, Supporting Information).43 Both atomically-flat basal planes
were {001} facets terminated with Cd ions coordinated with alkanoate ligands. To ensure the growth
of 2D nanocrystals via lateral extension, Dubertret et al. suggested that the size of seeds should be
identical to (or similar with) the thickness of the resulting 2D nanocrystals, i.e., 1.5 nm for 5.5-
23,28
monolayer CdSe 2D nanocrystals.
We noticed that this value was significantly smaller than the
seeds observed under transmission electron microscope (TEM) (see example in Figure S1, Supporting
Information), i.e., 1.8-2.2 nm in diameter. These results suggested that lateral extension through
addition of monomers might not be the growth mechanism of the 2D nanocrystals.
UV-Vis spectrum (Figure 1b) of the as-synthesized 2D nanocrystals using the purified seeds through
the scheme in Figure 1a was found to resemble that of high-purity 5.5-monolayer 2D nanocrystals
obtained in the original synthesis (Figure S2, Supporting Information). In comparison, the UV-Vis
spectrum of the purified seeds in Figure 1b was typical for CdSe quantum dots. The absorption peak
ACS Paragon Plus Environment

Page 7 of 32
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
was found to be consistent with that of dot-shaped CdSe 0D nanocrystals with the size being ~1.8 nm.
44 The TEM measurements revealed that the top view of 2D nanocrystals synthesized using the purified
seeds (Figure 1c) was nearly identical to that of the 2D nanocrystals synthesized with the unpurified
seeds (Figure S2, Supporting Information). Specifically, the average dimensions of the width and
length were ~8 and ~45 nm (Figure S3, Supporting Information), respectively. As shown in the high-
resolution TEM images and corresponding selected-area FFT (fast Fourier transform) patterns (Figures
1e and S4 in Supporting Information), two groups of d-spacing along <100> and <110> directions
could be observed, which identified the thickness direction (perpendicular to the view plane under
those TEM images) and lateral directions in Figure 1d and 1e. Specifically, the side facets of the
rectangles were the same as the basal planes, i.e., {100}. Though all thickness, width, and length
directions were <100> of the zinc-blende structure, their dimensions were significantly different from
each other (Figures 1d, 1e, S4 (Supporting Information)).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Results discussed above clearly indicated that growth of the CdSe 2D nanocrystals, specifically the
symmetry-breaking between thickness, width, and length directions, did not require existence of
cadmium and selenium precursors. Thus, these preliminary results suggested that oriented attachment
should be a plausible mechanism for the transition from the nanocrystal seeds to the 2D nanocrystals.
Ligands on growth of the CdSe 2D nanocrystals. Oriented attachment implies critical importance
of the surface ligands of nanocrystals. It was reported that, upon reduction of ligand protection, several
types of nanocrystal systems would become unstable and form complex structures by oriented
36-40
attachment.
For the current system, while cadmium acetate (Cd(Ac)2) and cadmium stearate
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 32
1
2
3
4
5
6
7
8
9
(Cd(St)2) were both in the solution, introduction of the purified seeds into the solution could generate
2D nanocrystals with high purity at 250 oC (Figure 2a). Cd(St)2 alone would keep the seeds stable in
the growth solution, but Cd(Ac)2 would drastically destabilize the seeds (Figure 2b). Some of the seeds
remained unchanged with the Cd(Ac)2-only reaction
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
while relatively large and irregular species were
observed in the UV-Vis and PL measurements (Figure
2b). The minor sharp feature at ~550 nm in the PL
spectrum of the resulting nanocrystals of the Cd(Ac)2-
only reaction, and TEM image in Figure S5
(Supporting Information) showed that there are a small
fraction of the 2D nanocrystals in the mixed products.
It was reported that fatty amines were necessary
Figure 2. With the same (purified) CdSe
ingredients for formation of CdSe 2D nanocrystals in
seeds, UV-Vis and PL spectra of CdSe
3,22,32,45
nanocrystals prepared with the existence of
wurtzite
structure.
Conversely,
(a) both cadmium acetate (Cd(Ac)2) and
cadmium stearate (Cd(St)2), (b) either
Cd(Ac)2 or Cd(St)2, (c) Cd(Ac)2 mixed with
organophosphine ligands were found to hinder the
either
oleylamine
(NH2Ol),
or
growth of CdSe 2D nanosheets in zinc-blende
tributylphosphine
trioctylphosphine
(TBP),
oxide
46
(TOPO),
structure.
Those two types of ligands and
respectively. For clarity, PL spectra are not
shown in (c).
organophoshpine oxides—another type of common
ligands—were tested in the reactions with the purified
seeds. Results in Figure 2c revealed that none of these three types of ligands would efficiently promote
formation of 2D nanocrystals by mixing them with Cd(Ac)2. When phosphine oxide and Cd(Ac)2 were
ACS Paragon Plus Environment

Page 9 of 32
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
in the solution, no sign of the 2D nanocrystals was detected. The other two types of ligands seemed to
promote random aggregation caused by Cd(Ac)2, indicated by complete disappearance of the original
seeds (compare Figures 2c with 2b).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Free fatty acids were found to be somewhat similar to the corresponding cadmium fatty acid salts
(Figure S6, Supporting Information). Under elevated temperatures in octadecene, stearic acid would
partially replace acetate in Cd(Ac)2 and the resulting salts with mixed carboxylate groups would
become soluble. Presumably, this would convert the growth solution to be similar with the typical
growth system in Figure 1a.
Replacement of cadmium acetate by acetic acid for typical reaction was not effective to promote
formation
of
the
2D
nanocrystals (data not shown).
Presumably, the boiling point
of acetic acid was too low to
keep it in the reaction solution
under the necessary high
reaction temperatures.
Figure 3. (a) Relative CdSe-unit yield of 2D nanocrystals versus the
concentration of Cd(Ac)2, with a fixed amount of Cd(St)2. (b) UV-Vis
spectra of the final products for three representative reactions in (a). (c)
Relative CdSe-unit yield of 2D nanocrystals versus the concentration of
Cd(St)2, with a fixed amount of Cd(Ac)2. (d) UV-Vis spectra of the final
products for three representative reactions in (c).
Different roles of cadmium
acetate
and
cadmium
stearate. Results in the above
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 10 of 32
1
2
3
4
5
6
7
8
9
sub-section suggested that both cadmium acetate and cadmium stearate (or other n-alkanoates with a
long hydrocarbon chain, see Figure S7, Supporting Information) were necessary for conversion of the
seeds to the 2D nanocrystals. However, additional results revealed that these two reagents likely played
different roles (Figure 3).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3a reveals that, at a fixed concentration of cadmium stearate, the final yield of the 2D
nanocrystals for each reaction increased rapidly in the beginning and reached a plateau as the
concentration of cadmium acetate increased. The CdSe-unit yield of the 2D nanocrystals was
represented by the absorbance at 550 nm. This was so because the molar extinction coefficients at the
peak position of the CdSe 2D nanocrystals with a given thickness, i.e., 550 nm for 5.5-monolayer CdSe
2D nanocrystals, were reported to be approximately linearly related to their area of the basal planes. 47
Evidently, the low conversion yield of the reactions with limited amounts of cadmium acetate was
accompanied with the remaining of the 0D nanocrystals (Figure 3b).
For the reactions with a fixed concentration of cadmium acetate, the final yields of the 2D nanocrystals
increased rapidly to a peak and followed by a steady decrease as the concentration of cadmium stearate
increased (Figure 3c). Figure 3d shows that no 0D nanocrystals were left at the peak of conversion.
However, on both sides of the conversion peak, existence of 0D nanocrystals was clearly evidenced.
By varying the hydrocarbon chain length of n-alkanotes from 14 to 22, the general trend in Figure 3
remained. Interestingly, though no 0D seeds left in the solution, the final yield of 2D nanocrystals
increased steadily as the hydrocarbon chain length increased (Figure S7, Supporting Information),
ACS Paragon Plus Environment

Page 11 of 32
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
indicating existence of some not fully developed nanostructures (see further discussion below).
Furthermore, the longer the hydrocarbon chain of the n-alkanotes, the slower the initial conversion rate
was observed (Figure S7, Supporting Information).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The results in Figures 2, 3, S6, and S7 suggested that cadmium n-alkanotes with a long hydrocarbon
chain played two roles for the formation of 2D nanocrystals. Firstly, it assisted cadmium acetate to
convert the 0D nanocrystals to 2D nanocrystals. In the current reaction solution, cadmium acetate was
found to be insoluble. It was confirmed that the existence of cadmium n-alkanotes with a long
hydrocarbon chain would greatly improve the solubility of cadmium acetate by formation of cadmium
salts with mixed carboxylate groups. Secondly, it stabilized the 0D nanocrystal seeds. The first role
was dominating at low concentrations of the cadmium n-alkanotes with a long hydrocarbon chain,
while the second one became more and more important when their concentration passed the conversion
peak.
Conversion process from 0D seeds to 2D nanocrystals. With optimized concentrations, the
conversion process was studied. Figure 4 illustrates the temporal evolution of UV-Vis/PL spectra (left
panel) and the morphology (right panel) of the nanocrystals for a reaction at 250 oC.
Quantitative results of the UV-Vis spectra are summarized in Figure S8 (Supporting Information),
which determined that the absorbance at 320 nm during the entire reaction course remained nearly
constant. This means that a constant CdSe unit yield in the form of nanocrystals throughout the entire
conversion process. 48 Consistent with this conclusion, a constant yield of CdSe units was also obtained
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 12 of 32
1
2
3
4
5
6
7
8
9
by using the extinction coefficients of the 0D seeds and 2D nanocrystal final products (Table S1,
Supporting Information). A constant yield of CdSe units in the form of nanocrystals excluded growth
from free precursors in solution, in accordance with the effective purification mentioned above.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Alternatively, instead of oriented attachment, a constant CdSe-unit yield during the conversion reaction
might be explained by a special pathway, i.e., a very small portion of the seeds growing rapidly by
slow dissolution of the other seeds. Against this hypothesis, the size evolution of 0D nanocrystals
followed an opposite trend,
i.e., rapid increasing in size
within the first ~30 s and then
remaining at a nearly constant
size (see Figure 4 left and
discussions below).
During the first 30 s of the
reaction in Figure 4, the UV-
Vis peak shifted from 480 nm
Figure 4. Temporal evolutions of UV-Vis/PL (left) and TEM images (right)
of CdSe nanocrystals at 250 oC. The dashed line (left panel) indicates the
to 490 nm, and the PL peak
absorption peak position of the persistent 0D nanocrystals.
evolved from a symmetric and
narrow peak to a peak with shoulders on both sides. During the entire conversion process, the
absorption peak at ~490 nm was observable except at the very end of the reaction (see the UV-Vis
spectrum at 1200 s). Under TEM, the size of the nanocrystals in the first 30 s was found to increase
ACS Paragon Plus Environment

Page 13 of 32
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
slightly to ~2.0 nm from ~1.8 nm seeds, and their two-dimensional projection on TEM remained
approximately spherical (Figure S1, Supporting Information). As the reaction proceeded beyond 300
seconds, population of tiny 0D nanocrystals decreased and gradually became scarcely observable
(Figure 4, right panel).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The results described in the above paragraph implied that the 0D nanocrystals with their first excitonic
absorption peak at ~490 nm were the critical intermediates for attachment. Given their nearly spherical
shape and similar sizes of the seeds, we named them as “single-dot intermediates” for attachment. To
confirm existence of such special intermediates, the 0D seeds with their first excitonic absorption peak
in the range between 470 to 520 nm were examined (Figure S9, Supporting Information). All those
reactions demonstrated formation of the single-dot intermediates during the early stage of a conversion
reaction, which had the first excitonic absorption peak at 490 ± 2 nm. Interestingly, for the 0D seeds
with their first excitonic absorption peak at ~490 nm, formation of the single-dot intermediates from
the seeds was still needed prior to formation of the 2D nanocrystals (Figure S9, Supporting
Information). Furthermore, for those reactions with different seeds, it was possible to identify similar
transient stages with nearly identical UV-Vis and PL spectra, i.e., with both absorption features for the
single-dot intermediates and 2D nanocrystals (Figure S9, Supporting Information).
The TEM images in Figure 4 (right panel) revealed that the lateral dimensions of 2D nanocrystals
increased gradually and the relative population of the single-dot intermediates decreased. Though the
distribution of lateral dimensions was not monodispersed at all stages, it gradually narrowed by the
time nearly all small particles—presumably the single-dot intermediates—were completely consumed.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 14 of 32
1
2
3
4
5
6
7
8
9
The lateral dimensions of the 2D nanocrystals in the early stage (before 120 s for the reaction in Figure
4) was mostly irregular in shape and smaller than the final values. During the middle stage (between
~180 and ~300 s), the width for the large ones in the solution was almost fully developed but the length
was noticeably smaller than the final value.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Interestingly, two or three side facets of the large 2D nanocrystals in the middle stage were quite well
developed as {100} facets but one or two of their side facet(s) remained as {110} ones. Their
interesting morphology was well resolved by Scanning Transmission Electron Microscope and high-
resolution TEM (Figure S10, Supporting information). The polar {100} facets of zinc-blende CdSe
nanocrystals should be terminated with cadmium ions that would be well coordinated with carboxylate
ligands. 33 Conversely, ligands passivation on the nonpolar {110} facets would be comparatively poor.
Thus, these large 2D nanocrystals in the middle stage should grow by particle attachment onto the
active {110} facet(s), which would eventually result in 2D nanocrystals in rectangular shape with four
{100} side facets. It was interesting to notice that formation of PbS nanosheets was also identified
through oriented attachment of the {110} facets of the 0D nanocrystals. 35
The PL spectra in Figure 4 (left panel) revealed additional evidences for particle attachment, instead
of growth through lateral extension by monomer addition. In principle, the smallest volume increase
for particle attachment should be one nanocrystal (the precursor) and that for lateral extension by
monomer addition should be one atom/molecule. The former would show a significant peak jumping
in the evolution of PL spectra, and the PL peak of the latter should shift continuously. Evidently, the
experimental results in Figure 4 (left panel) was consistent with that of particle attachment. Four PL
ACS Paragon Plus Environment

Page 15 of 32
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
peaks dominated all PL spectra in Figure 4 (left panel), namely ~490 nm for the seeds, ~505 nm for
the single-dot intermediates, ~538nm for another intermediate species (to be discussed below), and
550 nm for the 2D nanocrystals. These four peaks appeared sequentially. As mentioned above, the PL
spectra of all reactions with different size seeds were qualitatively the same after disappearance of the
original seeds (Figure S9, Supporting Information). The PL peak of the seeds disappeared after ~1 min
of the reaction, and that for the single-dot intermediates lasted roughly from 30 s to nearly the end of
the conversion, which was consistent with the UV-Vis spectra described above. The PL peak for the
2D nanocrystals at 550 nm became more and more dominating along the reaction. The peak at ~538
nm became evident shortly after the appearance of the ~505 nm peak associated with the single-dot
intermediates, and lasted to the end of a reaction though the intensity became very faint after ~600 s
for the reaction in Figure 4. Furthermore, the PL peak at ~538 nm (varied between 530 nm to 540 nm
for different reactions) was always observed regardless of size of the seeds used.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
All the results in the above paragraph not only supported the particle-attachment mechanism but also
suggested some structural information about the species related to the PL peak at ~538 nm. These
species should be associated with the smallest ones grown by particle attachment of the single-dot
intermediates. Thus, we would denote them as “2D embryos” and shall devote an entire subsection to
study their structure.
The spectral features related to Figure 4 and described above were observed for all reactions with
different cadmium n-alkanoates and/or different reaction temperatures. For the reactions with different
cadmium carboxylate concentrations, the features in Figure 4 were also observed once the reaction
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 16 of 32
1
2
3
4
5
6
7
8
9
resulted in significant conversion. Further experiments revealed that, for the original synthesis without
purification of the seeds, the temporal evolution of optical properties and shape of the nanocrystals
was quite similar to those in Figure 4 (Figure S11, Supporting Information). All these results suggest
a universal mechanism for formation of the 2D nanocrystals with absorption peak at 550 nm from the
0D seeds.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Temperature effects on formation of 2D nanocrystals. Results in Figure 5(a) illustrate the temporal
evolution of the conversion yield—the CdSe unit yields in 2D nanocrystals—of the reactions at
different reaction temperatures. As pointed out above, this conversion yield would not include the 2D
embryos because it was measured by the absorbance at 550 nm of reaction aliquots. Evidently, the
conversion yield increased rapidly by increasing the reaction temperature. The initial inductive time
approximately corresponded to formation of the single-dot intermediates from the seeds, which
decreased
increase
with
the
of
reaction
temperature.
All curves in Figure 5(a)
evidently did not follow
mono-exponential
Figure 5. (a) The relative conversion yield of CdSe units from 0D to 2D
nanocrystals under different temperatures. (b) Arrhenius plot for the
conversion from the single-dot intermediates to the 2D nanocrystals with
t1 as the time to reach a fixed CdSe-unit yield in 2D nanocrystals. (c)
Arrhenius plot for the conversion from the seeds to the single-dot
intermediates with t0 as the inductive time in Figure 5(a).
increase, indicating
a
complex kinetics involved
in the formation of 2D
ACS Paragon Plus Environment

Page 17 of 32
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
nanocrystals from the single-dot intermediates. This seemed to be reasonable, considering the
complicated chemistry involved in oriented attachment. Nevertheless, one could obtain experimental
activation energy (Ea) by Arrhenius plot as a semi-quantitative indicator of the temperature-
dependence of two processes, namely formation of the single-dot intermediates from the seeds as well
as conversion of the single-dot intermediates to the 2D nanocrystals. The rate for the former case at a
given temperature was calculated as the inverse of the inductive time in Figure 5(a), and the rate for
the latter case was estimated as the inverse of reaction time to reach a fixed CdSe-unit yield of the 2D
nanocrystals given in Figure 5(a). The Ea values in Figures 5(b) and 5(c) imply that conversion of the
single-dot intermediates to 2D nanocrystals should be substantially less costly in energy, though both
processes needed significant thermal activation. It should be noted that the single-component fitting in
Figure 5, especially in Figure 5b, is not perfect, which should only be applied for determining the
temperature-dependence of the related process and does not exclude complex chemical kinetics for the
corresponding process.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
It is interesting to notice that formation of CdSe wurtzite 2D nanocrystals occurred under relative low
o
3,22,32,45
temperatures, i.e., between 60-140 C.
Based on the results in Figure 5, neither formation of
the single-dot intermediates from the seeds nor conversion of the single-dot intermediates to the 2D
nanocrystals could be observed if the reaction was below 140 oC within the experimental timeframe.
The substantially low reaction temperatures for formation of wurtzite 2D nanocrystals should be
related to two factors. Firstly, the seeds for the wurtzite CdSe 2D nanocrystals were believed to be tiny
“magic size clusters”, which are thermodynamically less stable than the seeds and single-dot
intermediates in this work. Secondly, formation of the wurtzite CdSe 2D nanocrystals was suggested
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 18 of 32
1
2
3
4
5
6
7
to rely on the soft templates composed of alkylamine-carboxylate layer structures, which would
become unstable at elevated temperatures.
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Characterization of the 2D embryos. Results above suggested that the conversion of seeds to 2D
nanocrystals went through two intermediates with reasonably defined structures, i.e., the single-dot
intermediates and 2D embryos. In this sub-section, we shall discuss the 2D embryos, which should
help us to illustrate the structure of single-dot intermediates in the following sub-section.
As discussed above, attachment of the single-dot intermediates lasted until the end of the conversion
process. The 2D embryos might thus exist towards the end of the conversion process.
Though the
population would
be very low, the
2D
would likely be
able to be
embryos
separated from the
2D nanocrystals,
given their very
distinctive sizes.
Figure 6. (a) UV-Vis spectra of supernatant and precipitate separated by
centrifugation of aliquots of the final solution. (b) high resolution TEM image with
a few representatives {110} and {100} facets labeled. Tthe FFT pattern was
obtained from all nanocrystals. (c) UV-Vis spectra of the seeds, single-dot
intermediates, and 2D embryos. (d) Overlapping comparison of absorption features
between single-dot intermediates and 2D embryos after aligning the main peak.
By
high-speed
centrifugation of
the final product
ACS Paragon Plus Environment

Page 19 of 32
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
of a reaction (see Experimental for detail), we obtained a large amount of materials as precipitate and
a light-color supernatant. The precipitate was re-dispersed into solution and its UV-Vis was found to
be consistent with high-purity 2D nanocrystals (Figure 6a). In comparison to that of the aliquots before
separation, the UV-Vis spectrum of the re-dispersed precipitate showed a reduced absorbance at ~528
nm and missed two minor peaks at ~410 and 429 nm (Figure 6a). Though the supernatant had a very
small amount of nanocrystalline materials, it was sufficient for recording UV-Vis spectrum and TEM
measurements.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The UV-Vis of the 2D embryos in the supernatant, with its main absorption peak at ~528 nm, was
found to be qualitatively different from that of the 2D nanocrystals (Figure 6a). Furthermore, the two
missing minor peaks at ~410 and 429 nm in UV-Vis spectrum of the re-dispersed precipitate versus
that of the original aliquots were fully recovered in the spectrum of the supernatant. The tail to the low
energy side of the main peak was likely associated with the presence of a minor amount of the random
aggregates caused by cadmium acetate (see Figure 2b).
TEM measurements (Figure S1, Supporting Information) identified that the 2D embryos were
significantly larger than the single-dot intermediates. For the reaction using the 1.8 nm seeds (UV-Vis
peak at 480 nm), the average size of the single-dot intermediates and the 2D embryo was ~2.0 nm and
~3.0 nm (Figure S1, Supporting Information), respectively. Assuming the dimension perpendicular to
the TEM substrate was similar for single-dot intermediates and the 2D embryos, one would conclude
that one 2D embryo should be produced by two single-dot intermediates.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 20 of 32
1
2
3
4
5
High-resolution TEM images of the 2D embryos (Figure 6b) revealed that nearly all nanocrystals
6
7
49
possessed a clear lattice fringes, which is unusual for spherical dots in zinc-blende structure.
8
9
Furthermore, the zone axis of the ensemble in Figure 6b was identified as <100>. Furthermore, the
average size of 3 nm for the 2D embryos was found to be significantly larger than a typical dot with
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
44
the same absorption peak position. These facts strongly imply a possible morphology for the 2D
embryos, i.e., thin disks with their basal planes as well-developed {100} facets.
While the basal planes should be well developed and atomically flat, the top view of these 2D embryos
was mostly irregular (Figure 6b). These were thought to be consistent with oriented attachment of two
single-dot intermediates. Consistent with flat yet relatively large basal planes and thin thickness, the
lattice fringes of nearly all nanocrystals in Figure 6b were single-crystalline. Majority of the side facets
could be identified as {100} and {110} (see representative ones in Figure 6b). Given the high reactivity
of the {110} facets versus the {100} facets mentioned above, such types of side facets of the 2D
embryos were found to be reasonable. Both their formation from two single-dot intermediates and their
further growth into 2D nanocrystals should have gone through oriented attachment at the {110} side
facets.
Structure of single-dot intermediates. The single-dot intermediates appeared slightly larger than the
typical seeds under TEM, specifically ~2.0 nm for the single-dot intermediates versus ~1.8 nm for the
typical seeds (Figure S1, Supporting Information). As mentioned above, assuming the same dimension
along the direction perpendicular to the TEM substrates, each 2D embryo (Figure 6b) was
approximately equivalent to two single-dot intermediates in total CdSe units.
ACS Paragon Plus Environment

Page 21 of 32
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Figure 6c illustrates that the single-dot intermediates possessed a set of distinguishable absorption
features (~490 nm, ~441 nm, ~392 nm, and ~374 nm). While the main absorption peaks in the UV-Vis
spectra for the seeds and single-dot intermediates were close in energy, the spectral contour of the
single-dot intermediates was found to be distinguishable from that of the seeds. Instead, the contour of
UV-Vis spectrum for the single-dot intermediates was found to be quite similar to that of the 2D
embryos (Figure 6c). In fact, the UV-Vis absorption spectra and fluorescence excitation spectra of the
single-dot intermediates and 2D embryos could almost overlap with each other by aligning their first
excitonic peaks (see Figures 6d and S12 in Supporting Information). Specifically, both single-dot
intermediates and 2D embryos possessed a pair of absorption peaks separated by ~18 nm, namely
~392/374 nm for the single-dot intermediates and ~410/427 nm for the 2D embryos. These spectral
features along with dimensions estimated by TEM suggested that the single-dot intermediates might
likely possess 2D-like structure, meaning with well-developed {100} facets as atomically flat basal
planes and thin dimension perpendicular to the basal planes. To confirm this hypothesis, an early-stage
(60 s) aliquots—presumably enriched in single-dot intermediates as shown in Figure 4—was studied
by high-resolution TEM.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Results in Figure S13 (Supporting Information) demonstrated that, similar to Figure 6b, nearly all
nanocrystals of this early-stage aliquots possessed clear lattice fringes and the zone axis of the
ensemble was also <100>. In addition to a few of those relatively large nanocrystals—supposedly the
2D embryos and other 2D nanocrystals, there were quite some small nanocrystals (~2 nm in size) in
Figure S13 (Supporting Information). The zone axis of these small nanocrystals was <100> and the
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 22 of 32
1
2
3
4
5
6
7
side facets included both {100} and {110} ones, which was found to be consistent with the structure
of single-dot intermediates discussed in the above paragraph.
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Results discussed above implied that formation of the single-dot intermediates from the seeds was
likely intra-particle ripening. 8 Through such an intra-particle ripening process, a typical seed (1.8 nm
in all three dimensions) shrank one dimension to the thickness of a 2D nanocrystal (1.5 nm) and
expanded the other two directions to ~2.0 nm. At the same time, two atomically flat {100} basal planes
were developed. Such surface reconstruction was not completely surprising, given the high
concentration of cadmium carboxylates in the reaction solution. It was observed that, with a high
concentration of cadmium carboxylates, CdSe nanocrystals in zinc-blende structure preferred to be
faceted with exposure of their low index facets (mostly {100} and {110}, and minor contributions of
{111}). 49 Presumably, the {110} side facets could survive in the single-dot intermediates (Figure S13,
Supporting Information) and the 2D embryos (Figure 6b), but would be gradually eliminated in the
final 2D nanocrystals (Figure 1) by either oriented attachment or intra-particle ripening.
Formation mechanism of the 2D nanocrystals from the seeds. With all results disclosed in this
report, it should be possible to conclude that 5.5-monolayer CdSe 2D nanocrystals in zinc-blende
structure were formed through a complex mechanism involving oriented attachment. The conversion
of the 0D seeds to the 2D nanocrystals involved several key steps and two symmetry-breaking events.
In the first step, seeds with the sizes somewhat larger than the thickness of the targeting 2D
nanocrystals—between 1.7 and 2.2 nm applied in this work for 5.5-monolayer CdSe 2D nanocrystals
ACS Paragon Plus Environment

Page 23 of 32
Journal of the American Chemical Society
1
2
3
4
with 1.5 nm thickness—would adjust their shape through intra-particle ripening at elevated
5
6
7
o
temperatures (> ~200 C). This would result in single-dot intermediates with nearly flat {100} basal
8
9
planes and thickness similar to that of the final 2D nanocrystals (1.5 nm for 5.5-monolayer CdSe 2D
nanocrystals). The shrinkage along the thickness dimension would slightly expand the lateral
dimensions of the single-dot intermediates (Figure S1, Supporting Information), and their side facets
include stable {100} ones and other reactive ones, such as {110} facets (Figure S13, Supporting
Information). Formation of this type of single-dot intermediates required existence of extra cadmium
carboxylates. Cadmium fatty acid salts with a long hydrocarbon chain were the critical reagent for this
step (Figure 3, right). Cadmium acetate would accelerate this step but could induce uncontrolled
aggregation without existence of the long-chain alkanoates (Figure 3, left).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In the second step, two single-dot intermediates fused together through oriented attachment to form
one 2D embryo. This step was substantially slower than the first step (Figure 4). Because the {100}
facets, especially the atomically flat basal planes, were well passivated with carboxylate ligands, the
oriented attachment occurred at the reactive side facets of the single-dot intermediates, such as {110}
ones. Such oriented attachment continues between the single-dot intermediates and 2D embryos (or
other reactive 2D nanocrystals), resulting in irregular lateral dimensions in the early stage of oriented
attachment (Figure 4, right panel). This would complete the first symmetry breaking on the basis of
the first step, which differentiated the thickness direction—one of three <100> direction—and the two
lateral directions. The Weller group also suggested a similar facet-selective oriented attachment for
PbS nanocrystals in rock salt structure, which named the {110} facets as the attachment front for the
35
face-center-cubic lattice. Different from their proposed mechanism, structure characterization of
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 24 of 32
1
2
3
4
5
6
7
8
9
CdSe 2D embryos and single-dot intermediates in Figures 6 and S13 (Supporting Information) strongly
suggested that the CdSe nanocrystals developed their atomically flat basal planes at least in the form
of 2D embryos, possibly in the form of the single-dot intermediates. A key factor for completing the
first symmetry-breaking at this step was the relatively fast formation of single-dot intermediates and
very slow oriented attachment, which was ~60 s for the first step and > 600 s for the second step as
revealed by Figure 4. Such rate difference between two processes could be achieved as long as the
reaction temperature was not excessively high, given their significantly different activation energies
(Figure 5). A slow second step further allowed oriented attachment to occur with one single-dot
intermediate at one time and leave sufficient time for adjusting the side facets in the third step.
Different from the first step, oriented attachment in this step as well in the next step would require
small carboxylates, specifically acetates, on the surface of the nanocrystals to ensure insufficient-
ligand-protection through rapid ligand dynamics. 38,39,50,51 Because of insolubility of cadmium acetate
in the solvent, cadmium alkanoates with a long hydrocarbon chain were needed to act as shuttles to
bring acetates onto the surface of nanocrystals (Figures 2 and 3).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In the third step, single-dot intermediates continuously attached onto the 2D embryos as well as other
reactive 2D nanocrystals with irregular lateral shape. Simultaneously, the resulting 2D nanocrystals
adjusted their side facets to {100} ones through intra-particle ripening (Figure 4, right panel). Possibly,
oriented attachment between 2D embryos and other small 2D nanocrystals might also occur in this
step, but it should have not played a major role because the shape and lateral dimensions of the final
2D nanocrystals were reasonably monodisperse. Similar to the second step, oriented attachment should
occur on the {110} or other active side facets, which would retain the first symmetry-breaking achieved
ACS Paragon Plus Environment

Page 25 of 32
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
in the second step. The remaining {110} and other reactive side facets of the resulting 2D nanocrystals
would gradually be eliminated through intra-particle ripening, which should be somewhat similar to
that proposed by the Weller group for formation of PbS nanosheets. 35 Such an intra-particle ripening
process would gradually start to define the top view of the 2D nanocrystals when their lateral
dimensions reached ~8 nm in the current reaction system. Not surprisingly, the slow intra-particle
ripening did not develop four {100} side facets simultaneously, which would cause the second
symmetry-breaking (see 180 s TEM in Figure 4 and high-resolution TEM in Figure S10, Supporting
Information). The oriented attachment barely occurred onto three well-developed {100} side facets
and continued onto the remaining {110} side facet(s), which would result in significant elongation
along one of two lateral dimensions (Figures 1 and 4).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The first and second steps had minimum overlapping in the time domain but the second and third step
overlapped for the majority period of a reaction. As a reaction proceeded, the second step became less
and less competitive because concentration of the single-dot intermediates decreased rapidly. At the
same time, the concentration and the reactive surface area of the 2D nanocrystals increased
significantly to promote the third step. As a result, the width and length of the 2D nanocrystals were
reasonably monodispersed.
Conclusion
In conclusion, for zinc-blende CdSe (face-center-cubic), it is possible to grow 2D nanocrystals encased
with six {100} facets with distinctive dimensions of thickness, width, and length through a complex
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 26 of 32
1
2
3
4
5
6
7
8
9
mechanism involving oriented attachment. By applying purified seeds with their sizes between 1.7 and
2.2 nm for growth of 5.5-monolayer CdSe 2D nanocrystals (1.5 nm in thickness), growth through
lateral extension by monomer addition onto existing seeds was excluded. The first symmetry breaking
(between the thickness and lateral directions) was realized by rapid formation of single-dot
intermediates through intra-particle ripening of the seeds, which possessed a reduced thickness (1.5
nm) perpendicular to the atomically flat {100} basal planes and expanded lateral dimensions in
comparison to diameter of the initial seeds. The second symmetry breaking (between two lateral
dimensions) originated from unequal development of the polar and well passivated {100} side facets
during the oriented attachment. Cadmium acetate assisted both formation of single-dot intermediates
and oriented attachment. Cadmium alkanoates with a long hydrocarbon chain also played two roles in
the conversion of the seeds to 2D nanocrystals, namely, selectively stabilizing polar {100} facets of
the nanocrystals and shuttling insoluble acetate to the surface of the nanocrystals. Preliminary results
implied that oriented-attachment mechanism also worked for other types of zinc-blende cadmium
chalcogenide 2D nanocrystals (data not shown). Understanding symmetry breaking in highly
symmetric crystal structures should help to open a door for synthesis of a diverse group of 2D
nanocrystals with unique optical and optoelectronic properties.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Experimental Section:
Materials. Cadmium acetate dihydrate (Cd(Ac)2·2H2O, 99.999%), selenium powder (Se, 200 mesh,
99.999%), stearic acid (HSt, 90+%), oleic acid (HOl, 90%), butylamine (98%), octanoic acid (HOc,
98+%), and 1-octadecene (ODE, 90%) were purchased from Alfa-Aesar. Oleylamine (NH2Ol, C18
ACS Paragon Plus Environment

Page 27 of 32
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
content 70%) was purchased from Aldrich. Tributylphosphine (TBP) was purchased from Shanghai
Titan Chem. Tri-n-octylphosphine oxide (TOPO) was purchased from J&K. Acetone (analytical
reagent), methanol absolute and chloroform (analytical reagent) were purchased from Sinopharm
Reagents. All chemicals were used without further purification.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Synthesis of cadmium stearate (Cd(St)2). Typically, 20 mmol of HSt and 20 mmol of
tetramethylammonium hydroxide were dissolved in 200 ml of methanol. To this solution,
Cd(Ac)2·2H2O (10 mmol) dissolved in 50 ml of methanol was added dropwise under vigorous stirring.
White Cd(St)2 then precipitated from the reaction mixture. After the addition was completed, the
mixture was kept on stirring for another 20 min. The white precipitate was separated by centrifugation,
washed three times with methanol, and dried under vacuum overnight for usage.
Synthesis of other types of cadmium fatty acid salts was carried out using the same procedure described
above. Stearic acid was replaced by an equal molar amount of the corresponding fatty acid.
Direct synthesis of 5.5-monolayer CdSe 2D nanocrystals. In a typical synthesis, Cd(St)2 (0.2 mmol)
and ODE (3 ml) were loaded in a 25 ml three-neck flask. After degassing for 10 min, the mixture in
the flask was heated to 250 oC under Ar protection. A total amount of 8 mg (0.1 mmol) Se dispersed
in 1 mL ODE were swiftly injected. When the lowest energy absorption peak of the product in the UV-
Vis spectrum reached at ~490 nm, 40 mg Cd(Ac)2·2H2O (0.15 mmol) was introduced. The reaction
was kept at 250oC for several minutes and aliquots were taken out for monitoring the reaction by UV-
Vis. The reaction was stopped by allowing the reaction mixture to cool down in air.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 28 of 32
1
2
3
4
5
6
7
8
9
Synthesis and purification of CdSe nanocrystals (seeds) of three different sizes (the first excitonic
absorption peak at 473 nm, 488 nm, 502 nm, respectively). Atypical synthesis of CdSe nanocrystals
was conducted by injecting a 2.0-mLSe-octadecene suspension (0.1 mol/L) into a hot (260 oC) mixture
of Cd(St)2 (0.2714 g, 0.4 mmol) and 13 mL ODE in a 50-mL three-neck flask. The reaction was kept
at 240oC for several minutes. Needle-tip aliquots were taken for UV–vis and PL measurements to
monitor the reaction. When the targeted size of nanocrystals was achieved, the reaction mixture was
cooled rapidly to room temperature for purification using a procedure in literature.41
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The reaction mixture was centrifuged at 12000 rpm for 10 min, and the precipitate was dissolved in a
mixture of chloroform (4 ml) and butylamine (60 μl), and then kept at 60 °C to form a clear solution.
Acetonitrile with equivalent volume to chloroform/butylamine mixture was added into the vial. The
vial was centrifugated at 4000 rpm for ∼1 min, while the solution temperature was still at ∼50 °C. The
supernatant was quickly removed, with precipitate dissolved in ∼1.0 mL chloroform at 60 °C . 1.2 mL
acetonitrile was then added into it again to start another cycle of precipitation, centrifugation, and
decantation. The final products were re-dispersed in 2 ml ODE at 65°C , and low boiling point residuals
were removed from the solution by Ar bubbling at 65 °C .
Synthesis of CdSe 2D nanocrystals using purified CdSe nanocrystal seeds. Typically, Cd(St)2
(variable amount) and ODE were loaded in a 25 ml three-neck flask. After bubbling for 10 min, the
mixture in the flask was heated to a designated temperature (250 oC as the default) under Ar protection.
Pre-crushed Cd(Ac)2 suspended in 1 mL ODE was injected into the mixture, followed by the stock
ACS Paragon Plus Environment

Page 29 of 32
Journal of the American Chemical Society
1
2
3
4
5
6
solution of CdSe nanocrystal seeds. The amount of seeds was usually the same as that used in the direct
synthesis of 2D nanocrystals without purification of the seeds. The reaction temperature was set at a
7
8
9
o
designated value (250 C as the default) and the total volume of solution was controlled at 5 ml.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Needle-tip aliquots were taken and dispersed in 0.2% oleic acid (or octanoic acid) solution in toluene
for UV–vis and PL measurements to monitor the reaction. When the reaction was completed, several
drops of oleic acid were added to the flask prior to the cooling of mixture.
Purification and separation of CdSe 2D nanocrystals and 2D embryos. The finial product of a
reaction was centrifugated at 10000 rpm for 5 min. The precipitate was obtained by decantation and
dissolved in chloroform. Acetonitrile was added and the precipitate was isolated by centrifugation at
12000 rpm for 3 min. The precipitate containing CdSe 2D nanocrystals and supernatant with the 2D
embryos were isolated from each other by decantation. To improve the purity of either 2D nanocrystals
or 2D embryos, the cycle of dissolution, precipitation, centrifugation, and decantation might be
repeated multiple times for the designated component.
Measurements. UV-vis spectra were taken on Agilent Technologies Cary 4000 UV-Vis
spectrophotometer. PL spectra were recorded by Edinburgh Instruments FLS920 spectrometer. TEM
images were taken on a Hitachi 7700 transmission electron microscope operating at 100 kV, and the
nanocrystals were deposited onto a copper grid with ultrathin carbon film. HRTEM images were taken
on a JEM 2100F transmission electron microscope operating at 300 kV.
Supporting Information
ACS Paragon Plus Environment